CN107092579A - One kind is based on the improved SDFT frequency estimating methods of FFB - Google Patents
One kind is based on the improved SDFT frequency estimating methods of FFB Download PDFInfo
- Publication number
- CN107092579A CN107092579A CN201710164152.9A CN201710164152A CN107092579A CN 107092579 A CN107092579 A CN 107092579A CN 201710164152 A CN201710164152 A CN 201710164152A CN 107092579 A CN107092579 A CN 107092579A
- Authority
- CN
- China
- Prior art keywords
- ffb
- frequency
- signal
- current time
- sdft
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/141—Discrete Fourier transforms
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Discrete Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Complex Calculations (AREA)
Abstract
The improved SDFT frequency estimating methods of FFB are based on the invention discloses one kind, are comprised the following steps:1st, the sample frequency f of signal to be observed is determinedsWith the port number N of fast electric-wave filter group, and to observation signal with frequency fsProgress, which is sampled, obtains discrete observation sequence;2nd, the single pass parameters of FFB are determined, the fast electric-wave filter group of N channel is designed, obtains FFB equivalent coefficient h (n) and equivalence filter coefficient length L;3rd, by sequence inputting to be observed into FFB, then after the L+2 moment, FFB passage 1 has three output samples, is designated as X1、X0、X‑1;4th, formula is passed throughCalculate the frequency estimation at current time5th, get after the frequency estimation at current time, the X that current time is collected1,X0,X‑1Shifted, X is arrived in the output storage of the FFB passages 1 of subsequent time1In, go to step (4).It the method increase the mean square error performance of Frequency Estimation, and signal-noise ratio threshold.
Description
Technical field
The present invention relates to the signal frequency tracer technique of signal processing technology field, more particularly to power system.
Background technology
Frequency Estimation has extensively in fields such as communication, instrument, medical treatment, Speech processing, power system, etection theories
Using.
In field of power, system frequency is a key parameter for controlling electric power networks load unbalanced, can be determined
Harmonic current interference that nonlinear load is caused etc..Therefore, accurate frequency estimating methods are to maintaining electric power networks and conventional electricity
The stability of power equipment is most important.Grid equipment voltage signal is considered as pure string signal, then during the zero passage twice of the signal
Between interval be system frequency an important references.In fact, measured signal is easily disturbed by noise or other signals, it is existing many
The method that domestic and international researcher proposes a variety of signal frequencies estimations being directed under different scenes, including phase-lock-loop algorithm are planted,
Trapper, least mean square algorithm and its variant, sef-adapting filter, Kalman filter, Taylor expansion serial algorithm and iteration
Algorithm etc..
Frequency estimating methods based on discrete Fourier transform (Discrete Fourier Transform, DFT) are in electricity
It is widely used in Force system, but DFT causes that only in sample frequency be the whole of actual frequency to the sampling nature of frequency spectrum
When several times, just can accurately it be estimated.In fact, the shake of the data or signal frequency of finite length all can be to estimating
Meter produces spectral leakage and " fence effect ".In order to reduce these adverse effects, the DFT of power system signal resolution is improved
Rate, many methods are reduced " fence effect " by difference arithmetic and brought usually using spectral leakage is reduced based on window function
Error.However, difference arithmetic usually requires more DFT computings, therefore larger operand limits these algorithms in reality
Use in the scene of border.Much DFT performance is improved in addition, also having about the algorithm of asynchronous-sampling.For example based on instantaneous phase
Correction algorithm, the sampling rate adjusting of position error create condition of synchronized sampling etc..Also some algorithms use digital filter
The phase concussion of decay DFT components.A G Phadke et al. extract signal using the recurrence relation of two continuous DFT components
Instantaneous phase and system frequency, however, when system frequency and standard frequency have deviation, the hypothesis of synchronized sampling will reduce calculation
Method performance.In order to solve this problem, J Z Yang propose that one kind exists using the relation between the DFT fundamental components of continuous sampling point
Accurate algorithm for estimating under the scene of higher hamonic wave and system frequency deviation, and compare this algorithm (Smart Discrete
Fourier Transform, SDFT) performance advantage with the traditional sampling point algorithms of DFT tri- and Prony algorithms.In noisy field
The precision estimated under scape using SDFT algorithms frequency tends not to meet actual demand.
The content of the invention
Goal of the invention:For problems of the prior art, fast electric-wave filter group is based on the invention discloses one kind
(Fast Filter Bank, FFB) improved SDFT frequency estimating methods, the method increase the SDFT algorithms under noisy scene
The performance estimated frequency.
Technical scheme:One kind is based on the improved SDFT frequency estimating methods of FFB, comprises the following steps:
(1) the sample frequency f of signal to be observed is determinedsWith the port number N of fast electric-wave filter group, and to observation signal with frequency
Rate fsProgress, which is sampled, obtains discrete observation sequence;
(2) the single pass parameters of FFB are determined, the fast electric-wave filter group of N channel is designed, obtains FFB equivalent coefficient h (n)
With equivalence filter coefficient length L, wherein L=2 Γ+1, Γ represents h (n) the central point moment;
(3) by sequence inputting to be observed into FFB, then after the L+2 moment, FFB passage 1 has three output samples, note
For X1、X0、X-1, respectively current time, previous moment, the output at preceding two moment;
(4) formula is passed throughCalculate the frequency estimation at current timeWherein To take real part computing;
(5) get after the frequency estimation at current time, the X that current time is collected1,X0,X-1Shifted,
I.e. by X-1Cast out, X0Value storage arrive X-1, X1Value storage arrive X0, the output of the FFB passages 1 of subsequent time, which is stored, arrives X1In, turn
To step (4).
In order to carry out Frequency Estimation, sample frequency f to the signal of power systemsValue is Nf0, wherein N is fast electric-wave filter
Group FFB port number, f0For the standard frequency of power system signal, f0=50Hz.
Beneficial effect:Compared with prior art, frequency estimating methods disclosed by the invention overcome pair of FFT computings in itself
The damping capacity of secondary lobe is not enough, to noise-sensitive the shortcomings of, stronger FFB side lobe attenuation ability and noiseproof feature are incorporated into
In SDFT algorithms, the mean square error performance of Frequency Estimation, and signal-noise ratio threshold are substantially increased.
Brief description of the drawings
Fig. 1 is the frequency tracking result in the case where signal to noise ratio is -5dB scenes;
Fig. 2 is the frequency tracking result in the case where signal to noise ratio is 0dB scenes;
Fig. 3 is the frequency tracking result in the case where signal to noise ratio is 10dB scenes;
Fig. 4 is the frequency tracking result in the case where signal to noise ratio is 20dB scenes;
Fig. 5 is the inclined estimated bias curve of the SDFT algorithms based on FFT and FFB computings;
Fig. 6 is the inclined estimation mean square error curve of the SDFT algorithms based on FFT and FFB computings.
Embodiment
With reference to the accompanying drawings and detailed description, the present invention is furture elucidated.
One kind is based on the improved SDFT frequency estimating methods of FFB, comprises the following steps:
Step 1, the sample frequency f for determining signal to be observedsWith the port number N of fast electric-wave filter group, and to observation signal
With frequency fsProgress, which is sampled, obtains discrete observation sequence;
In order to carry out sample frequency f in Frequency Estimation, the present embodiment to the signal of power systemsValue is Nf0, wherein N is
Fast electric-wave filter group FFB port number, f0For the standard frequency of power system signal, f0=50Hz.
Step 2, determine the single pass parameters of FFB, design the fast electric-wave filter group of N channel, obtain FFB equivalent coefficient h
(n) with equivalence filter coefficient length L, wherein L=2 Γ+1, Γ represents h (n) the central point moment;
Step 3, by sequence inputting to be observed into FFB, then after the L+2 moment, FFB passage 1 has three output samples,
It is designated as X1、X0、X-1, respectively current time, previous moment, the output at preceding two moment;
Step 4, pass through formulaCalculate the power system signal for obtaining current time
Frequency estimationWherein To take real part computing, fs=Nf0;
Step 5, get after the frequency estimation at current time, the X that current time is collected1,X0,X-1Moved
Position, i.e., by X-1Cast out, X0Value storage arrive X-1, X1Value storage arrive X0, the output of the FFB passages 1 of subsequent time, which is stored, arrives X1
In, go to step (4).
So far, once complete frequency tracking process has been completed, over time, and constantly circulation carries out this and estimated
Meter process is the frequency tracking result that can obtain a period.
Formula in step 4Derivation it is as follows:
If without the power system signal u (t) that makes an uproar with frequency fsThe obtained discrete series v (n) of sample is:
V (n)=Acos (w0n△T+φ) (1)
In formula, A is power system signal amplitude, and φ is initial phase, and numerical frequency is w0=2 π f0, f0For power system
The standard frequency of signal, f0=50Hz.Signal sampling rate is△ T are the sampling interval, and N is the logical of FFB outputs
Road number, signal v (n) can be expressed as:
V=Ae in formulajφ, V*Represent V conjugation.Due to only considering the fundamental component of signal, it corresponds to the defeated of FFB passages 1
Go out, therefore do not consider the output of other channels.Output results of the v (n) after FFB passage 1 can be expressed as:
If considering system frequency deviation, ω0=2 π (f0+ △ f), substitutes into formula (2) and (3) and abbreviation can be obtained:
Order
The base of natural Exponents is further abbreviated as r, i.e.,:
Therefore, it can be obtained by above formula:
ak=rak-1,bk=r-1bk-1 (7)
The FFB outputs at continuous three moment are represented by
Above-mentioned three formula is done into simple arithmetical operation, can be obtained
In formula,Convolution (9) can further obtain f estimate:
To take real part computing, fs=Nf0,In formula,X is corresponded to respectively-1、X0、
X1。
Fig. 1-Fig. 4 is respectively to sinusoidal signal v (n)=Acos (w0N △ T+ φ) add -5dB, 0dB, 10dB, 20dB respectively
Noise, the frequency of signal is 51Hz, according to said frequencies tracing step, and the Frequency Estimation result calculated each moment is entered
Row is preserved.Fig. 1-Fig. 4 saves the result of calculation in 0.5s.
Fig. 5-Fig. 6 signal actual frequency is still 51Hz, for each snr value, carries out 100 above-mentioned computings,
The average and variance of this 100 operation results is taken to obtain respectively.
Claims (2)
1. one kind is based on the improved SDFT frequency estimating methods of FFB, it is characterised in that comprise the following steps:
(1) the sample frequency f of signal to be observed is determinedsWith the port number N of fast electric-wave filter group, and to observation signal with frequency fs
Progress, which is sampled, obtains discrete observation sequence;
(2) the single pass parameters of FFB are determined, the fast electric-wave filter group of N channel is designed, obtains FFB equivalent coefficient h (n) and wait
Filter coefficient length L is imitated, wherein L=2 Γ+1, Γ represents h (n) the central point moment;
(3) by sequence inputting to be observed into FFB, then after the L+2 moment, FFB passage 1 has three output samples, is designated as X1、
X0、X-1, respectively current time, previous moment, the output at preceding two moment;
(4) formula is passed throughCalculate the frequency estimation at current timeWherein To take real part computing;
(5) get after the frequency estimation at current time, the X that current time is collected1,X0,X-1Shifted, will
X-1Cast out, X0Value storage arrive X-1, X1Value storage arrive X0, the output of the FFB passages 1 of subsequent time, which is stored, arrives X1In, go to step
Suddenly (4).
2. according to claim 1 be based on the improved SDFT frequency estimating methods of FFB, it is characterised in that fsValue is Nf0,
Wherein N is the port number of fast electric-wave filter group, f0For the standard frequency of power system signal, f0=50Hz.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710164152.9A CN107092579A (en) | 2017-03-20 | 2017-03-20 | One kind is based on the improved SDFT frequency estimating methods of FFB |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710164152.9A CN107092579A (en) | 2017-03-20 | 2017-03-20 | One kind is based on the improved SDFT frequency estimating methods of FFB |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107092579A true CN107092579A (en) | 2017-08-25 |
Family
ID=59648893
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710164152.9A Pending CN107092579A (en) | 2017-03-20 | 2017-03-20 | One kind is based on the improved SDFT frequency estimating methods of FFB |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107092579A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107085140A (en) * | 2017-04-25 | 2017-08-22 | 东南大学 | Nonequilibrium system frequency estimating methods based on improved SmartDFT algorithms |
CN110674456A (en) * | 2019-09-26 | 2020-01-10 | 电子科技大学 | Time-frequency conversion method of signal acquisition system |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104202273A (en) * | 2014-09-15 | 2014-12-10 | 东南大学 | Phase-based frequency estimation interpolation direction judgment method |
CN106053936A (en) * | 2016-06-17 | 2016-10-26 | 海南大学 | Method and system for acquiring instantaneous frequency of electrical signal |
-
2017
- 2017-03-20 CN CN201710164152.9A patent/CN107092579A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104202273A (en) * | 2014-09-15 | 2014-12-10 | 东南大学 | Phase-based frequency estimation interpolation direction judgment method |
CN106053936A (en) * | 2016-06-17 | 2016-10-26 | 海南大学 | Method and system for acquiring instantaneous frequency of electrical signal |
Non-Patent Citations (3)
Title |
---|
JUN-ZHE YANG ET AL.: "A Precise Calculation of Power System Frequency and Phasor", 《IEEE TRANSACTIONS ON POWER DELIVERY》 * |
JUN-ZHE YANG ET AL.: "A Smart Method Makes DFT More Precise for Power System Frequency Estimation", 《CONFERENCE:POWER ENGINEERING SOCIETY 1999 WINTER MEETING,IEEE》 * |
张贵生: "基于多相滤波器组的频谱感知技术的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑(月刊)》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107085140A (en) * | 2017-04-25 | 2017-08-22 | 东南大学 | Nonequilibrium system frequency estimating methods based on improved SmartDFT algorithms |
CN107085140B (en) * | 2017-04-25 | 2019-04-16 | 东南大学 | Nonequilibrium system frequency estimating methods based on improved SmartDFT algorithm |
CN110674456A (en) * | 2019-09-26 | 2020-01-10 | 电子科技大学 | Time-frequency conversion method of signal acquisition system |
CN110674456B (en) * | 2019-09-26 | 2022-11-22 | 电子科技大学 | Time-frequency conversion method of signal acquisition system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107478990B (en) | A kind of generator electromechanical transient process method for dynamic estimation | |
CN110333389A (en) | Sinusoidal signal frequency estimation method based on interpolated DFT | |
CN109669072B (en) | Self-adaptive synchronous phasor measurement method for power distribution network | |
CN107092579A (en) | One kind is based on the improved SDFT frequency estimating methods of FFB | |
CN105137181A (en) | Double-spectrum-line interpolation harmonic analysis algorithm based on Nuttall-Kaiser composite window | |
Li | A total least squares enhanced smart DFT technique for frequency estimation of unbalanced three-phase power systems | |
CN108254794B (en) | Magnetic resonance denoising method and device based on modeling reverse recovery technology | |
CN108646091B (en) | Method for separating multi-component polynomial phase signals | |
CN111948454B (en) | Synchronous phasor measurement method and system for resisting attenuated direct current component interference | |
CN112883318A (en) | Multi-frequency attenuation signal parameter estimation algorithm of subtraction strategy | |
Hwang | Improvement of phasor estimation accuracy by prony-based identification of two decaying dc components | |
CN103605904A (en) | Self-compensating amplitude calculating method applied to electrical power system and based on error estimation | |
Hwang et al. | Identification of interarea modes from an effectual impulse response of ringdown frequency data | |
Subudhi et al. | A comparative study on different power system frequency estimation techniques | |
CN108173259B (en) | Sine frequency estimation method based on unit constraint minimum mean square error | |
CN113191317B (en) | Signal envelope extraction method and device based on pole construction low-pass filter | |
Rosolowski et al. | Fast removal of decaying offset in relaying input signals | |
CN115291151B (en) | High-precision magnetic resonance signal frequency measurement method based on low correlation segmentation | |
Jiang et al. | De-noising for partial discharge signals of power transformers using wavelet threshold method based on ant colony optimization | |
Sen et al. | Frequency estimation using nonlinear least squares and S-Transform | |
Rosolowski | A New Method for Fast Removal of Decaying Offset in Relaying Input Current | |
CN111079569B (en) | Spindle unbalanced vibration signal processing method based on full-phase FFT method | |
CN105468880B (en) | A kind of extracting method of low-frequency oscillation rapid decay signal component parameter | |
CN108020719A (en) | A kind of harmonic detecting method based on improvement windows and interpolated FFT | |
Duda et al. | Low-Complexity P Class to M Class PMU Model for Distribution Systems |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170825 |